Interview Experience – Amazon (Data Engineer Role)

My first round was a mix of DSA and basic concepts. They presented four questions:

• Valid Parentheses
• Sort 0s, 1s, and 2s
• Maximum Sum Subarray
• Reverse Words in a String

I solved three with optimal solutions and managed one with a brute-force approach (I didn't know the Dutch National Flag trick back then). Next, we covered DBMS concepts and a few Linux commands. I was comfortable with the basics but struggled with grep/sed-level commands. Honestly, I didn't expect to pass this round, but to my surprise, I did!

Round two intensified the challenge. They focused on SQL queries, specifically GROUP BY, aggregates, joins, and related concepts. I aced the first two queries, provided a partial solution for a difficult one, and then smoothly handled DBMS concepts like schema design, OLTP vs. OLAP, data warehousing, and data modeling. The final task involved a real-life database design scenario, which I actually enjoyed explaining and felt very confident about.

By the end, I genuinely thought I had nailed the interview. However, the reality check was that I wasn't selected. It turned out they only had one opening, which went to the very first candidate from my college. Life can be a little cruel sometimes, but that's how it goes. Still, it was a great learning experience. If you are preparing, make sure you are solid on DSA (don't skip optimal tricks!), SQL joins/aggregates, DBMS, and a bit of Linux.

DSA:

• Valid Parentheses
• Sort 0s, 1s, and 2s (Dutch National Flag problem)
• Maximum Sum Subarray (Kadane’s Algorithm)
• Reverse Words in a String

SQL / DBMS:

• Write queries using GROUP BY and aggregate functions
• Solve a complex JOIN query mixing multiple tables
• Conceptual questions on schema design, OLTP vs OLAP, data warehousing, and data modeling
• Scenario-based question: Design a real-life database schema and explain how you’d model and query it

Linux:

• Basics like navigating directories and file handling
• Missed out on tougher ones like grep and sed